distribution of heavy metals and arsenic in soft...

33

DISTRIBUTION OF HEAVY METALS ANDARSENIC IN SOFT SURFACE SEDIMENTS OF

THE COASTAL AREA OFF KOTKA,NORTHEASTERN GULF OF FINLAND,

BALTIC SEA

byHenry Vallius¹, Daria Ryabchuk ² and Aarno Kotilainen¹

Vallius, H., Ryabchuk, D. & Kotilainen, A. 2007. Distribution of heavymetals and arsenic in soft surface sediments of the coastal area off Kotka,North-Eastern Gulf of Finland, Baltic Sea. Geological Survey of Finland,Special Paper 45, 33–48, 12 figures and 5 tables.

The modern soft surface sediments of the sea area off Kotka in theNortheastern Gulf of Finland were surveyed for heavy metal contamina-tion. Altogether, 14 sites were sampled with a gravity corer for chemicalanalyses. The sea-floor sediments of this rather shallow and thus sensiti-ve sea area have for decades been loaded with heavy metals and otherharmful substances. Although a slightly improving trend can be seen, thesoft sediments off Kotka and Hamina can still be classified as largelypolluted, especially on the basis of cadmium and mercury contents in thesurface sediments. Deeper down, at depths of 10–20 cm, the sedimentsare even more highly contaminated, such that concentrations occasional-ly reach values representing very high contamination. Mercury is themain pollutant of the near coastal areas close to the outlets of River Kymi-joki, while cadmium is the main pollutant in the remote basins in the eastand southeast.

Key words (Georef Thesaurus AGI): Marine sediments, clay, mud,geochemistry, heavy metals, background level, Finland, Kotka, BalticSea, Gulf of Finland

¹ Geological Survey of Finland (GTK),P.O. Box 96, FI- 02151, Espoo, Finland

² A. P. Karpinsky Russian Geological Research Institute (VSEGEI),74 Sredny prospect, 199106, St.Petersburg, Russia

E-mail: ¹[email protected], ²[email protected]

Holocene sedimentary environment and sediment geochemistryof the Eastern Gulf of Finland, Baltic SeaEdited by Henry ValliusDistribution of heavy metals and arsenic in soft surface sediments ofthe coastal area off Kotka, North-Eastern Gulf of Finland, Baltic SeaGeological Survey of Finland, Special Paper 45, 33–48, 2007.

34

Geological Survey of Finland, Special Paper 45Henry Vallius, Daria Ryabchuk and Aarno Kotilainen

INTRODUCTION

the River Neva and the Neva Bay are more easilyforced towards the northeast, i.e. to the area of thisstudy. Here, the waters are mixed with those fromthe Vyborg Bay area and from the River Kymijoki.Thus, the geochemistry of the bottom sediments inthe study area is a mixture of components of localnatural origin combined with transported compo-nents from the north and east.

Eutrophication is a curse of the Baltic Sea (Pit-känen 1994 and Rantajärvi 1998) and it has alsostrongly influenced the sea in the study area, but asseveral authors have already reported on this prob-lem it will only be briefly mentioned in this paper.

This study was an outcome of the SAMAGOLproject, a collaborative project between the Geolog-ical Survey of Finland (GTK) and the All-RussiaGeological Research Institute (VSEGEI). The aimwas to gather available marine geological data fromthe eastern Gulf of Finland administered by the par-ticipating institutes in order to produce a marine geo-logical synthesis report from the study area. This pa-per describes the geochemistry of the recent surfacesediments in the study area, including arsenic (As),cadmium (Cd), cobalt (Co), chromium (Cr), copper(Cu), mercury (Hg), nickel (Ni), lead (Pb), vanadium(V), and zinc (Zn). Data presented in this paper arebased on samples that were collected in 2004.

The Gulf of Finland is an eastward extension ofthe Baltic Sea, an estuary-like and rather shallowsea. There is no tide and the sea is partly ice coveredduring the winter months. The salinity in the studyarea is low (2–4 salinity units) because of the ratherlimited water exchange with more saline waters inthe west and river run-off in the north and extremeeast of the Gulf. The total annual freshwater dis-charge to the Gulf of Finland is 112 km3 (Berg-ström & Carlsson 1994), of which most of it is dis-charged in the eastern Gulf of Finland, thus affect-ing the bottom in the study area. The Neva River isthe largest river in the area, with a mean annual run-off of some 2460 m3 s-1 (77.6 km3 a-1, Bergström &Carlsson 1994). The Neva River is, according toVallius & Leivuori (2003), the principal carrier ofpollutants, but the Vyborg Bay area and RiverKymijoki with its two outlets are also considerablesources of heavy metals. In addition, the cities ofHamina and Kotka, with rather large commercialharbours, add to the pollution of this sea area (Fig.1). The hydrography in the Gulf is controlled by theCoriolis force, which forces the currents into anti-clockwise movement (Palme’n 1930, Alenius &Myrberg 1998). This is further influenced by geo-morphology and meteorological factors. In the east-ern Gulf of Finland this means that currents from

Figure 1. Study area with the cities of Kotka and Hamina indicated. Arrows indicate the approximate location of the main outlets of River Kymijoki.Thickness of arrow indicates relative water flow. The dotted line is the border between Finland and Russia.

35

Geological Survey of Finland, Special Paper 45Distribution of heavy metals and arsenic in soft surface sediments of the coastal area off Kotka, North-Eastern Gulf of Finland, Baltic Sea

STUDY AREA

The study area covers both sides of the nationalborder between Finland and Russia, an area of ap-proximately 2000 km2 between the latitudes 60° 00and 60° 30 N and the longitudes 26° 30 and 28° 00 E(Fig. 1). It is situated on the northern coast of theGulf of Finland. The southern part of the area com-prises open sea with water depths of up to some 90metres. There are only a few bigger islands, ofwhich the largest is the island of Gogland, a horststructure reaching up to as much as 176 metresabove sea level. In the north, the study area bordersland and the archipelago is typically a mosaic of amultitude of islands differing in shape and size.Similarly, the sea floor in the north is made up of amosaic of rather small, relatively shallow and iso-lated accumulation basins. In the south the accumu-lation basins are clearly larger as the sea floor be-tween the sparse islands is more even. The waterdepth increases from land in the north to the deep-est parts in south. The deepest sampled site wasMGVV-2004-23, with a water depth of 72,0 metres.

Earlier studies

Several earlier studies have been published fromthe sea area around Kotka, Hamina and the outletsof River Kymijoki. This is mainly because the RiverKymijoki is known to be largely polluted by heavymetals and organic pollutants. According to Par-tanen (1993), the biota has largely been affected by

eutrophication in the outlets of River Kymijoki, butno chemical analyses were reported. A report byKokko & Turunen (1988) described considerablemercury contamination in the river sediment pro-file caused by large discharges during the late1950s and early 1960s. Anttila-Huhtinen & Heitto(1998), Verta et al. (1999a, b, and c) and Pallonen(2001, 2004) reported large-scale contamination ofriver and sea floor sediments of the River Kymijokiand sediments of the sea area outside the river out-lets. Leivuori (1998), Vallius & Lehto (1998), Val-lius & Leivuori (1999), Vallius (1999 a and b), Lei-vuori (2000) and Vallius & Leivuori (2003) havereported heavy metal distributions in the off-shoresediments of the Gulf of Finland, and in thesestudies the sea area outside Kotka and the RiverKymijoki has also been classified as significantly,largely or very largely contaminated, depending onthe element.

In several reports on the state of Finnish coastalwaters (National Board of Waters 1983, Pitkänen etal. 1987, and Kauppila & Bäck 2001), the sea areaoff the River Kymijoki has been classified as pol-luted or strongly polluted, depending on the sam-pling site. The most polluted sites are in the river,the river mouth, or immediately outside it. In allstudies mercury has been found to be the worst con-taminating heavy metal in the sediments. Cadmiumhas also been found in high or even very high con-centrations. In this study it was possible to comparethe results with the data from earlier studies.

Table 1. List of coring stations, coordinates in WGS-84.

Station Latitude Longitude Depth (m)

MGGN-04-10 60º29.255’N 27°07.275’E 15.6MGGN-04-13 60º20.141’N 27º35.485’E 39.0MGGN-04-14 60º24.069’N 26º57.141’E 20.2MGGN-04-15 60º11.298’N 27º06.029’E 63.5MGGN-04-16 60º29.940’N 27º45.774’E 16.0MGGN-04-17 60º23.816’N 27º39.434’E 43.5MGGN-04-18 60º28.111’N 27º30.370’E 14.3MGGN-04-19 60º23.043’N 27º27.252’E 30.0MGGN-04-20 60º29.537’N 27º03.062’E 13.2MGGN-04-21 60º23.117’N 27º17.723’E 38.3MGGN-04-22 60º23.053’N 27º16.577’E 38.0MGGN-04-23 60º10.751’N 27º05.432’E 72.0MGGN-04-24 60º19.428’N 26º52.788’E 38.0MGGN-04-25 60º24.644’N 27º36.320’E 39.7

MATERIALS AND METHODS

The samples of this study were recent muddyclays or clayey muds, which were collected duringthe SAMAGOL cruise of R/V Geola of the Geologi-cal Survey of Finland (GTK) on 24 May to 4 June2004. Altogether, 13 sites were sampled with agravity corer for chemical analyses. Additionally,one sample was taken onboard R/V Aranda inAugust 2004 (Table 1). The sampling sites were se-lected based on carefully surveyed echo-soundingdata. All samples were taken using a GEMAX twinbarrel gravity corer with an inner diameter of 90mm of the core liner. The core length varied be-tween 25 cm and 65 cm, and most of the coresreached 50 cm in length. After sampling, one of thetwo cores obtained was split vertically for descrip-tion and the other was used for sub-sampling. Thus,all cores were sliced into 10-mm-thick subsamples

36


onboard and packed into plastic bags that were im-mediately stored at –18 °C until they were takenashore for freeze drying (Leivuori 1998, Vallius &Lehto 1998, Vallius & Leivuori 1999 and 2003,Vallius 1999a,b, Leivuori 2000).

Chemical analyses were performed at the ana-lytical chemistry laboratory of GTK. The sampleswere completely digested with hydrofluoric – per-chloric acids followed by elemental determinationusing inductively-coupled plasma mass spectrome-try (ICP-MS) and inductively-coupled plasmaatomic emission spectrometry (ICP-AES) (Vallius& Leivuori 1999, 2003). Determination of C and Nwas performed using a LECO analyzer and mercurywas measured with an Hg analyzer through pyrolyticdetermination.

The analytical reliability of the laboratory waschecked using commercial standard reference mate-rials MESS-2 and NIST8704. All elements exceptcadmium had a recovery of ± 10% of the referencevalue, most of them within ± 5%, which can beconsidered as satisfactory or good. The average re-covery of cadmium was slightly too high forMESS-2 as it was 115%, but for NIST8704 it wasexactly 100%. It seems that the matrix of NIST8704(Buffalo River sediment) is more similar to the verylow salinity marine sediment of this study. In onesample batch the recovery of lead was 112% (aver-age of all batches for lead was 102%), while the re-covery for zinc in another batch was 89% (averageof all batches for zinc was 96%). Overall, the re-coveries of all studied elements were good enoughto be reliable.

In order to obtain a picture of the degree of con-tamination of the sediments in the study area theclassification of Vallius and Leivuori (2003) in theoffshore Gulf of Finland was used (Table 2). It isbased on the Swedish marine sediment quality cri-teria (Naturvårdsverket 1999, WGMS 2003), wheresurface concentrations are compared with back-ground values. The Swedish criteria are used because

no Finnish criteria are yet available. Quality crite-ria, which compare total concentrations with a refer-ence or with background values, provide little in-sight into the potential ecological impacts of sedi-ment contaminants, but they provide a base fromwhich to evaluate Sediment Quality Guidelines(SQG, Burton 2002). In this study the surface con-centrations as well as the concentrations at a depthof 14–15 cm have been classified in terms of thedegree of contamination.

The classification of contamination of the sur-face sediment layer gives a good picture of the con-dition of the sediment, but in order to evaluate fur-ther threats from the sediment column it is also nec-essary to look at the deeper layers. If the sedimentcolumn was for some reason disturbed, older sedi-ment from deeper layers would be exposed to thewater column. Dredging and other human activitiescan disturb the sediment column to great depths,but so too can bottom-dwelling animals such as thespionid polychaete Marenzelleria viridis, whichwas observed to burrow into the sediment to depthsof 15 cm or more. Thus, a classification of sedi-ment contamination at the depth of 14–15 cm in thestudied cores was added to the surface classifica-tion. The maximum contents of elements in differ-ent cores are dependent on the accumulation rates atdifferent depths. Thus, the depth of 14–15 cm doesnot correspond to the highest concentrations in allcores, but instead represents a depth, that can easilybe resuspended by human activity, bioturbation orcurrents.

The data are also presented as maps displayingthe horizontal distribution of the elements. As therewere too few samples for presentation as interpo-lated colour surface maps, the maps for arsenic,cadmium, cobalt, chromium, copper, mercury,nickel, lead, vanadium and zinc are presented as cir-cular symbol maps. Golden Software Surfer 8©was used in map production.

Table 2. Classification of sediment heavy metal contamination according to the Swedish Environmental Protection Agency (mg kg-1 dry weightbasis, total analysis, WGMS 2003). Classification for vanadium is missing.

Metal Class 1 Class 2 Class 3 Class 4 Class 5(mg kg-1) dry weight Little or none Slight Significant Large Very Large

As < 10 10–16 16–26 26–40 > 40Cd < 0.2 0.2–0.5 0.5–1.2 1.2–3 > 3Co < 14 14–20 20–28 28–40 > 40Cr < 80 80–110 110–160 160–220 > 220Cu < 15 15–30 30–60 60–120 > 120Hg < 0.04 0.04–0.10 0.10–0.27 0.27–0.7 > 0.7Ni < 33 33–43 43–56 56–80 > 80Pb < 31 31–46 46–68 68–100 > 100Zn < 85 85–125 125–195 195–300 > 300

37


RESULTS

& Lehto 1998, Leivuori 1998, Vallius & Leivuori1999, Leivuori 2000, and Leivuori & Vallius 2004).Samples from deeper basins had slightly higher ele-ment concentrations than those from shallowersites.

All studied surface samples showed a high de-gree of cadmium contamination (Fig. 3), the lowestmeasured concentration being 1.05 mg kg-1, whichstill is in the class of significant contamination(Table 2). At first it seems that cadmium concentra-tions increase with water depth, but at sites 15 and23, which are situated rather close to each other,with a distance of 1.1 km between them, the con-centrations differed markedly. Both sites are deep,with a water depth of 63.5 metres at site 15 (shownas a white circle on map) and a depth of 72.0 metresat site 23. However, the concentration of cadmiumat site 23 (2.69 mg kg-1) was twice as high as at site15 (1.21 mg kg-1) (Table 3). This difference canprobably be explained by the geological setting ofthese two sites. Site 15 is situated in the middle of alarge rather plane basin, while site 23 is situated onthe slope of a scar-like hole with steep slopes. Thus,site 23 represents an exceptional environment incomparison to the other sites, which are located inrather plane basins of differing shape and size.

When looking at the average trend for cadmium(Fig. 3), concentrations were clearly higher at theeastern and southeastern sites close to the nationalborder of Finland. It has been speculated (Vallius &Leivuori 1999) that due to geomorphological andhydrological factors, dissolved cadmium fromsource areas in the eastern Gulf of Finland is mostlytransported westward along the northern side of theGulf. Such transport from the east would explainthe higher concentrations of cadmium in the outerregions of the study area, as the off-shore currents

On the basis of earlier studies (Anttila-Huhtinen& Heitto 1998, Verta et al. 1999a,b, and c, Leivuori1998, Vallius & Lehto 1998, Vallius & Leivuori1999, Vallius 1999a, b, Leivuori 2000, Vallius &Leivuori 2003, and Pallonen 2001 and 2004) it wasexpected that the concentrations of certain elementswould be rather high in the sediments of the studyarea. Sedimentation rates were also expected to behigh. Thus, sampling aimed towards cores that wereas long as possible. Pre-industrial levels of heavymetals were reached in almost all cores. This depthvaried markedly depending on the sedimentationrate, but typically it was at a depth of about 40–50 cm if cadmium is used as indicator. At somesites, however, the maximum core length of 60 cmwas not enough to obtain a full sequence since pre-industrial times. At certain other sites, penetrationdeep into the sediment was hindered by the sedi-ment composition, with harder silty layers resultingin clearly shorter cores. However, all cores reacheda length of at least 20 cm, thus providing a goodrecord of heavy metal accumulation at the sampledsite.

Horizontal distribution of elements in thesoft surface sediments

The geochemical data of all studied elements inthe surface sediment (0–1 cm) are summarised inTable 3, while the maps show concentrations as cir-cular symbols with symbol size reflecting elementconcentration. The location of sampling sites is in-dicated in Figure 1.

All arsenic values in the study area were ratherlow (Fig. 2), which corresponds well with earlierstudies from the offshore Gulf of Finland (Vallius

Table 3. Surface sediment (0–1 cm) concentrations of studied elements. Concentrations in mg kg-1. For location of sampling sites see Figure 1.

Site Nr As Cd Co Cr Cu Hg Ni Pb V Zn Station

10 7.25 1.05 14.3 71.4 47.1 0.32 30.6 37.3 77.4 161 MGGN-04-1013 12.6 1.63 13.7 59.2 76.3 0.15 20.9 54.9 60.6 260 MGGN-04-1314 8.14 1.12 10.6 52.0 50.9 0.20 29.8 40.2 58.1 165 MGGN-04-1415 16.6 1.21 14.4 64.3 49.3 0.14 35.1 44.4 95.6 161 MGGN-04-1516 8.99 1.23 12.7 72.5 57.1 0.14 40.5 58.9 66.8 178 MGGN-04-1617 19.1 2.51 16.1 77.3 66.3 0.20 36.8 55.9 74.0 217 MGGN-04-1718 7.58 1.35 10.1 54.3 42.1 0.11 32.5 43.7 53.7 158 MGGN-04-1819 12.7 2.07 11.0 53.6 57.0 0.11 28.9 42.6 55.1 156 MGGN-04-1920 10.1 0.84 14.5 82.7 49.1 0.31 39.1 44.9 83.8 198 MGGN-04-2021 16.4 1.45 14.6 60.6 58.7 0.11 32.8 53.9 71.0 188 MGGN-04-2122 14.0 1.45 12.4 45.8 46.7 0.09 27.2 39.6 49.4 152 MGGN-04-2223 13.2 2.69 14.8 57.4 60.1 0.10 34.1 40.4 68.5 217 MGGN-04-2324 13.9 1.15 11.6 55.1 49.0 0.16 29.7 40.1 65.9 153 MGGN-04-2425 14.3 2.30 13.4 51.6 47.2 0.15 32.6 48.1 55.5 167 MGGN-04-25

38


Figure 2. Distribution of arsenic in the surface sediments (0–1 cm). Concentrations in mg kg-1.

Figure 3. Distribution of cadmium in the surface sediments (0–1 cm). Concentrations in mg kg-1.

39


do not enter the shallower near-coastal areas. Ac-cording to the data presented by Anttila-Huhtinen& Heitto (1998) and Pallonen (2001, 2004), thecadmium concentrations in the outer sea area to theeast and southeast of Kotka seem to have clearlydecreased from 1994 to 2003. To apply this idea tothe outermost sites of this study is, however, slight-ly speculative as no sites in the three above-men-tioned studies were situated as far off-shore as theoutermost samples of this study. Unfortunately, asnone of the sites were from exactly the same loca-tion, comparison between these studies is less reli-able.

Concentrations of both cobalt (Fig. 4) and chro-mium (Fig. 5) were rather low and equally distribut-ed. Mean concentrations of chromium (61.3 mg kg-1,Table 4) in this study were on a slightly lower levelthan the mean average concentrations of chro-mium in the studies of Anttila-Huhtinen & Heitto(1998) and Pallonen (2001, 2004). This might beattributed to a cleaner trend of release of thesemetals.

Copper concentrations were on average ratherhigh (Fig. 6). Average surface concentrations (mean

54.1 mg kg-1, Table 4) were of the same order ofmagnitude as the mean values recorded in earlierstudies by Anttila-Huhtinen & Heitto (1998) andPallonen (2001, 2004). At three sites (13, 17 and23) the concentrations exeeded 60 mg kg-1, with aconcentration of 76.3 mg kg-1 measured at site 13,66.3 mg kg-1 at site 17 and 60.1 mg kg-1 at site 23(Table 3). All these sites are located very close tothe border between Finland and Russia and a simi-lar transport to that speculated for cadmium mightbe responsible for these rather high concentrations.The location of these anomalous sites and the factthat copper concentrations at these sites were higherthan surface concentrations at any site during a sur-vey in the early 1990s by Vallius & Leivuori (1999)implies an increasing trend in copper release in theeasternmost part of the Gulf of Finland. Pallonen(2004) also noted an overall slight increase in cop-per concentrations from 1994 to 2003.

Average mercury concentrations (Fig. 7) werelower than the average recorded in earlier studies byAnttila-Huhtinen & Heitto (1998) and Pallonen(2001, 2004) in this area and by Vallius & Leivuori(1999) from the whole off-shore Gulf of Finland.

Figure 4. Distribution of cobalt in the surface sediments (0–1 cm). Concentrations in mg kg-1.

40


Figure 5. Distribution of chromium in the surface sediments (0–1 cm). Concentrations in mg kg-1.

Figure 6. Distribution of copper in the surface sediments (0–1 cm). Concentrations in mg kg-1.

41


The samples from sites 10 and 20 with concentra-tions of 0.32 and 0.31 mg kg-1, respectively, clearlydiffered from the average pattern. These sites arelocated rather close to the eastern outlet of the RiverKymijoki (Fig. 1) and thus the much higher concen-tration could be attributed to the well-known mer-cury load of the river (Kokko & Turunen 1988, Ant-tila-Huhtinen & Heitto 1998, Verta et al. 1999a,b,and c, and Pallonen 2001 and 2004). As samplenumber 17, taken close to the border, had a clearlyhigher concentration (0.20 mg kg-1) and its neigh-bouring sites 13 and 25 also had slightly higherconcentrations (0.15 mg kg-1), it could be speculat-ed that some of the mercury is transported from theeast, but this is rather speculative. Similarly, sam-ples 14 and 24 had higher mercury concentrations(0.16 and 20 mg kg-1, respectively), which could beattributed to their location closer to the main west-ern outlet of the River Kymijoki (Fig. 1).

The horizontal distribution of nickel in the sur-vey area (Fig. 8) was even, thus being similar to thedistribution of cobalt and chromium. No cleartrends could be found. Leivuori (2000) reportednickel values from 20 stations in the Gulf of Fin-

land that were slightly higher than those of thepresent study. Leivuori (2000) recorded a mean val-ue of 42 mg kg-1 (32.2 mg kg-1 in this study, Table 4)and minimum and maximum values of 25 mg kg-1

and 60 mg kg-1, respectively (20.9 mg kg-1 and40.5 mg kg-1 in this study). Thus, nickel concentra-tions in the coastal area seem to be somewhat lowerthan in the open sea area of the Gulf of Finland. Asthe data of Leivuori’s study were from the early1990s it is also possible that there has been a de-crease in nickel loading to the sediment since thattime. On the other hand, Pallonen (2004) stated thatin the Kotka area there were both increases and de-creases in nickel concentrations from 1994 to 2003,depending on site.

Lead (Fig. 9) exhibited a similar horizontal pat-tern to cobalt, chromium and nickel as the distribu-tion was even, with no clear trends in any direction.The concentrations of lead were clearly lower thanin the earlier study by Vallius & Leivuori (1999)from the off-sore Gulf of Finland in the early 1990s.That earlier study reported a mean lead value of 51mg kg-1 while the mean of this study was 46.1 mg kg-1

(Table 4). The maximum value of 58.9 mg kg-1 of

Figure 7. Distribution of mercury in the surface sediments (0–1 cm). Concentrations in mg kg-1.

42


Figure 8. Distribution of nickel in the surface sediments (0–1 cm). Concentrations in mg kg-1.

Figure 9. Distribution of lead in the surface sediments (0–1 cm). Concentrations in mg kg-1.

43


this study was much lower than the maximum valueof 88 mg kg-1 in the earlier study. Pallonen (2004)also reported a considerable decrease in lead con-centrations, in some case up to 20 to 40%, from2000 to 2003. In the case of lead it is probable thatthis difference can be attributed to a further reduc-tion in the lead load that was already observed inthe early 1990s (Vallius & Lehto 1998 and Vallius& Leivuori 1999).

The horizontal distribution of vanadium (Fig.10) did not exhibit any clear trends, as slightlyhigher concentrations were found in various loca-tions around the study area. Vanadium concentra-tions from 20 stations in the off-shore Gulf of Fin-land have earlier been reported by Leivuori (2000).The average values of the present study were slight-ly lower than those of that earlier study, but themaximum values of the two studies were on samelevel. Leivuori (2000) reported a mean value of 76mg kg-1 (66.8 mg kg-1 in this study, Table 4) andminimum and maximum values of 57 mg kg-1 and96 mg kg-1, respectively (49.4 mg kg-1 and 95.6mg kg-1 in this study, Table 4). The difference in

mean concentrations could be attributed to differ-ences in sample site location or to a real decrease inthe vanadium load.

The distribution of zinc showed only a veryslight elevation of zinc concentrations in the vicini-ty of the Finnish-Russian border (Fig. 11). Other-wise, the distribution was quite even throughout thestudy area. When comparing the data from thisstudy with the earlier study of Vallius & Leivuori(1999), the mean concentrations in the samples ofthis study (181 mg kg-1, Table 4) were slightly lowerthan those of the earlier study (199 mg kg-1). Themaximum concentration of 260 mg kg-1 zinc in thisstudy was also clearly lower than the maximum of391 mg kg-1 zinc recorded earlier. This difference isprobably due to differences in the location of sam-pling sites, with slightly different environmentssampled, as Pallonen (2004) noted that in the Kotkaarea the concentrations of zinc remained rather con-stant from 1994 to 2003. According to Pallonen(2004), there have been both increases and decreas-es in zinc concentrations in the coastal area, de-pending on the sample site.

Figure 10. Distribution of vanadium in the surface sediments (0–1 cm). Concentrations in mg kg-1.

44


Classification of contamination

Surface sediments (0–1 cm)

As shown in Table 4, most classified metalswere in the contamination classes of little or none(class 1) or slight (class 2) when examining the sur-face concentrations. Only cadmium, copper, zincand mercury were clearly strong contaminants ofthe seafloor sediments in the study area (Natur-vårdsverket 1999). All these metals were in theclasses of significant or large contamination, butthey were still below the values representing verylarge contamination (class 5).

Deeper sediments (14–15 cm)

Table 5 presents the concentrations of studiedmetals in the depth layer of 14–15 cm at the studiedsites. As expected, the concentrations were higherin the deeper layers and thus the sediment was inmany cases classified as more strongly contaminat-ed than in the surface sediments. The worst case

was mercury, whose maximum content of 0.83mg kg-1 reached the class of very large contamina-tion (class 5). This sample (MGGN-2004-20) wasnot surprisingly from a bay situated immediatelyoutside the eastern outlet of the River Kymijoki.The maximum cadmium concentration of 4.45mg kg-1 was also recorded at a site close to the RiverKymijoki outlet (MGGN-2004-10), but in anotheraccumulation basin closer to the port of Hamina;thus, the source of this anomalous peak probablyoriginated in the port instead of the river. This isalso supported by the fact that the site closest to theriver mouth (MGGN-2004-20) showed clearly low-er cadmium values. When looking at the mean val-ues for the elements it is clear that in addition tomercury and cadmium, copper, lead and zinc arealso strong contaminants of the sediments in thestudy area.

Vertical distribution

As can be seen in the classification of the degreeof contamination of the cores, there was a cleartrend of decreasing concentrations towards the

Figure 11. Distribution of zinc in the surface sediments (0–1 cm). Concentrations in mg kg-1.

45


surface of the sediment for all elements except ar-senic. Such a trend was also observed in the early1990s by Vallius & Lehto (1998), Vallius (1999),Vallius & Leivuori (1999) and Leivuori (2000). Asan example, Figure 12 illustrates mercury, cadmiumand zinc concentrations in cores 20 and 17, thesesites representing the extremes in terms of the con-centrations of these contaminants. Site 20 is locatedclose to the river mouth while site 17 is the mostdistantly situated site in the east (Fig. 1), very close

to the border of Finland and Russia. These curvesshow the typical trend of metal concentrations incores from the study area. The depths of the anoma-lies depend on the sampling site, mostly due to dif-ferences in the sedimentation rate, but the cores dis-play a common pattern of concentration changes.First, there is an increase from the bottom of thecore from pre-industrial values to maximum values,then an anomaly with rather high concentrations,and usually a slow decrease in concentrations near

Table 4. Concentrations and descriptive statistics for studied elements in the surface sediments (0–1 cm). Concentrations in mg kg-1. See the ex-planations of classes in Table 1. Classification for vanadium is missing in the Swedish EPA

As class Cd class Co class Cr class Cu class Ni class Pb class V class Zn class Hg class

Mean 12,5 2 1,56 4 13,2 1 61,3 1 54,1 3 32,2 1 46,1 2 66,8 181 3 0,16 3Median 13,0 2 1,45 4 13,6 1 58,3 1 50,1 3 32,6 1 44,1 2 66,4 166 3 0,15 3SD 3,64 0,58 1,80 10,9 9,18 5,02 7,06 12,9 31,8 0,07Minimum 7,25 1 0,84 3 10,1 1 45,8 1 42,1 3 20,9 1 37,3 2 49,4 152 3 0,09 2Maximum 19,1 3 2,69 4 16,1 2 82,7 2 76,3 4 40,5 2 58,9 3 95,6 260 4 0,32 4Count 14 14 14 14 14 14 14 14 14 14

Table 5. Concentrations and descriptive statistics for studied elements in the sediments at a depth of 14–15 cm. Concentrations in mg kg-1. Seethe explanations of classes in Table 1.

As class Cd class Co class Cr class Cu class Ni class Pb class V class Zn class Hg class

Mean 13,3 2 2,30 4 18,6 2 91,9 2 62,9 4 41,9 2 69,1 4 90,8 237 4 0,33 4Median 13,1 2 2,47 4 17,8 2 90,7 2 64,2 4 40,4 2 68,4 4 89,2 238 4 0,31 4SD 4,24 1,23 4,79 12,4 14,2 3,97 17,8 12,4 44,7 0,19Minimum 9,24 1 1,00 3 12,5 1 66,6 1 28,1 2 36,8 2 21,1 1 70,4 112 2 0,17 3Maximum 25,4 3 4,45 5 30,3 4 111 3 82,8 4 49,3 3 87,8 4 114 297 4 0,83 5Count 14 14 14 14 14 14 14 14 14 14

Figure 12. Vertical profiles of mercury (Hg), cadmium (Cd), and zinc (Zn) at stations 20 and 17. Concentrations in mg kg-1.

46


the surface. In core 20 the decrease in the mercuryload from maximum values in the late 1960s isclearly visible. In the same core, cadmium concen-trations are rather low in comparison to other sitesin the study area. Like mercury, zinc also showsrather high concentrations in the upper half of thecore, with only a slight decrease during the last de-cades. In core 17, mercury concentrations are clear-ly lower than at site 20 and the concentrations havebeen steadily decreasing. Cadmium, on the otherhand, shows high concentrations in the upper halfof the core, with a clear decrease during the last de-cade or two. The overall concentrations of cadmiumare clearly higher in core 17 than core 20. It is nota-ble that at the depth of 3–9 cm in core 17 the con-centrations exceed 3 mg kg-1, which is the threshold

for classification as having very large contamina-tion (Table 2), reaching a maximum value of 4.57mg kg-1 at the depth of only 5–6 cm. Thus, sedi-ments very close to the sediment surface at site 17are very strongly contaminated with cadmium. Zincin core 17 shows a similar pattern to that in core 20,with an almost negligible decrease during the lastdecades. Copper has followed a similar trend tozinc, but particularly at site 17 near the border itseems that copper concentrations have not de-creased during the last decades. This is consistentwith the finding of Pallonen (2004), who also notedan overall slight increase in copper concentrationsfrom 1994 to 2003. The findings at this site canprobably be attributed to transport from source ar-eas in the eastern part of the Gulf of Finland.

DISCUSSION

According to the findings in this study, the Kot-ka–Hamina sea area is in a broad sense a badly pol-luted area, perhaps one of the worst polluted in thewhole Baltic Sea. This is not a surprising result, assimilar findings have earlier been published fromthis area (Kokko & Turunen 1988, Anttila-Huhtinen& Heitto 1998, Verta et al. 1999a,b, and c, and Pal-lonen 2001 and 2004). Studies from the offshoreGulf of Finland have also indicated that not only theeastern part of the Gulf but also the northeasternpart between the cities of Vyborg in Russia andKotka are largely polluted (Vallius & Leivuori1999, Vallius 1999b, Leivuori 2000, and Vallius &Leivuori 2003).

The horizontal distribution of most metals doesnot follow any clear pattern, as there are slightanomalies in different parts of the study area, whileon the other hand mercury and cadmium displayrather distinct anomalies. Mercury is the main pol-lutant of the near-coastal areas close to the outlets ofthe River Kymijoki, while cadmium is the main pol-lutant in the remote basins in the east and southeast.

The vertical profiles indicate that concentrationsof most of the studied metals have decreased duringthe last decade or two, but the concentrations of es-pecially mercury, cadmium, copper and zinc arestill much too high in the sediment surface. As thereare even higher concentrations deeper in the sedi-ment, which could easily be exposed by human ac-tivity or bottom-dwelling animals and erosion, thesituation cannot be considered satisfactory. Copperconcentrations seem to decline rather slowly, and atsite 17 its concentration has not decreased at allduring the last decades.

This study has shown that the distribution ofcadmium in the sea area off Kotka is zoned, suchthat cadmium concentrations in the sediment sur-face increase towards the east and southeast. This isbetter seen in the maps of this study than in earlierstudies (Kokko & Turunen 1988, Anttila-Huhtinen& Heitto 1998, and Pallonen 2001 and 2004), as thestudy area here has been extended all the way to theFinnish–Russian border. The higher concentrationsclose to the border can probably be attributed tocadmium transport to the west from sources in theeasternmost Gulf of Finland. This might also be thecase for some of the other heavy metals, such ascopper and zinc.

The pollution load in the study area is generallydecreasing, which can also be seen in the sea-floorsediments. Unfortunately, the process is slow andthe concentrations have been so high that a longtime will be needed for the area to be normalized.When the surface sediments are classified as large-ly polluted, in many cases due to high concentra-tions of especially cadmium and mercury, the situa-tion cannot be considered satisfactory. As there areeven more highly polluted sediments close to thesediment surface, at depths of only 10–20 cm wherethe sediment can easily be exposed due to humanactivity or burrowing by bottom-dwelling animals,the situation is still very chronic. If dredging is car-ied out in any parts of the area, which is of coursenecessary at least along the shipping channels,more metals will be released from the deeper sedi-ments. This will also be the case if dredging is al-lowed in the river itself.

47


CONCLUSIONS

Geology also plays a role in the interpretation ofthe level of contamination. As the surveyed area islocated within the large Vyborg rapakivi massif, thenatural concentrations of many of the studied met-als in this area are lower that in areas of supracrustalrocks. Thus, the actual degree of contamination inthis area is probably even higher than the measuredlevel, as the classification does not take into ac-count cases where natural levels of metals are clear-ly lower than average.

The classification used in this paper is takenfrom Sweden, as no similar classifications areavailable in Finland. The problem with the Swedishclassification is that the intervals within each classseem to be too narrow. In addition, small differenc-es in element concentrations between separate stud-ies from different years might not be of relevance,as the small differences almost fall within the rangeof analytical accuracy. Thus, such findings shouldbe looked upon as only suggestive.

ACKNOWLEDGEMENTS

for their help during the cruises. Thanks are alsoexpressed to the staff of the Geolaboratory of theGTK (Labtium Oy since September 2007) . Finally,the authors wish to thank the directors and the fi-nancing agencies of both GTK and VSEGEI for theopportunity to realize the SAMAGOL project.

The team of VSEGEI with Dr. Mikhail Spiri-donov as leader was of great importance during thecruise of R/V Geola. As the scientific teams and thecrews of the vessels of the cruises were too big foreverybody to be mentioned, all scientists and thecrews of R/V Geola and R/V Aranda are thanked

REFERENCES

Alenius, P. & Myrberg, K. 1998. Hydrodynamics ofthe Gulf of Finland. In: Sarkkula (ed.) Proceedingsof the Final Seminar of the Gulf of Finland Year1996. Suomen Ympäristökeskuksen monistesarja105, 331–339.

Anttila-Huhtinen, M. & Heitto, L. 1998. Harmfulsubstances in Kymi River and its outlets, resultsfrom the 1994 study. Haitalliset aineet Kymijoella jasen edustan merialueella, tuloksia vuoden 1994 tut-kimuksista. Kymijoen vesiensuojeluyhdistys ry:njulkaisu 75. 83 p. (in Finnish)

Bergström, S. & Carlsson, B. 1994. River runoff to theBaltic Sea: 1950–1990. Ambio 23 (4–5), 280–287.

Burton, G. A. Jr. 2002. Sediment quality criteria in usearound the world. Limnology 3, 65–75.

Kauppila, P. & Bäck, S. (Eds) 2001. The state of Fin-nish coastal waters in the 1990s. The Finnish Envi-ronment 472. 134 p.

Kokko, H. & Turunen, T. 1988. Mercury loading intothe lower River Kymijoki and mercury concentra-tions in pike until 1986. Kymijoen alaosaan kohdis-tunut elohopeakuormitus ja hauen elohopeapitoisuusvuoteen 1986 saakka. Vesitalous 3, 30–38. (inFinnish)

Leivuori, M. 1998. Heavy metal contamination in sur-face sediments in the Gulf of Finland and compari-son with the Gulf of Bothnia. Chemosphere 36,43–59.

Leivuori, M. 2000. Distribution and accumulation ofmetals in sediments of the northern Baltic Sea. FinnishInstitute of Marine Research – Contributions 2. 159 p.

Leivuori, M. & Vallius, H. 2004. Arsenic in marine se-diments. In: Kirsti Loukola-Ruskeeniemi & PerttiLahermo (eds.) Arsenic in Finland: Distribution,Environmental Impacts and Risks. Geological Sur-vey of Finland, 89–96. (in Finnish with synopsis inEnglish)

National Board of Waters, 1983. State of the waters inearly 1980’s. Subreport # 7 of the Water ProtectionTarget Programme Vesistöjen tila 1980- luvun alus-sa. Vesiensuojelun tavoiteohjelmaprojektin osara-portti 7. National Board of Waters Report 194. 24 p.(in Finnish).

Naturvårdsverket 1999. Bedömningsgrunder för mil-jökvalitet- Kust och hav (Assessment of Environ-mental Quality- in Coast and Sea). Naturvårdsver-ket. Rapport No. 4914. 134 p. URL: www.internat.naturvardsverket.se/documents/legal/assess/assess.htm.Visited 11.05.2007.

Pallonen, R. 2001. Harmful substances in sediments inthe sea area off River Kymijoki in autumn 2000 Hai-talliset aineet Kymijoen edustan sedimenteissä syk-syllä 2000. Kymijoen vesiensuojeluyhdistys ry:njulkaisu 93. 19 p. (in Finnish).

Pallonen, R. 2004. Harmful substances in sediments inthe sea area off River Kymijoki in autumn 2003.

48


Haitalliset aineet Kymijoen edustan sedimenteissäsyksyllä 2003. Kymijoen vesi ja ympäristö ry:n jul-kaisu 112. 24 p. (in Finnish)

Partanen, P. 1993. Biota studies in Lake Konnivesi,River Kymijoki and off-shore Pyhtää, Kotka andHamina in 1992 (in Finnish). PohjaeläintutkimuksetKonnivedellä, Kymijoella sekä Pyhtään, Kotkan jaHaminan merialueilla v. 1992. Kymijoen vesiensuo-jeluyhdistys ry:n tiedonantoja 39. 39 p.

Palmén, E. 1930. Untersuchungen über die Strömun-gen in der Finnland umgebenden Meeren. Investiga-tions of currents in the sea areas surrounding Fin-land. Societas Scientiarum Fennica. Commentatio-nes Physico- Mathematicae V, 12. (In German)

Pitkänen, H, Kangas, P., Miettinen, V. & Ekholm, P.1987. The state of the Finnish coastal waters in1979–1983. National Board of Waters and the Envi-ronment, Finland. Publications of the Water and En-vironment Administration 8. 167 p.

Pitkänen, H, 1994. Eutrophication of the Finnish coastalwaters: Origin, fate and effects of riverine nutrientfluxes. National Board of Waters and the Environ-ment, Finland. Publications of the Water and Envi-ronment Research Institute 18. 44 p.

Rantajärvi, E. 1998. Leväkukintatilanne Suomen me-rialueilla ja varsinaisella Itämerellä vuonna 1997.Phytoplankton blooms in the Finnish Sea areas andin the Baltic Proper during 1997. Meri report seriesof the Finnish Institute of Marine Research 36, 32 p.(in Finnish with summary in English)

Vallius, H. 1999a. Anthropogenically derived heavymetals and arsenic in recent sediments of the Gulfof Finland, Baltic Sea. Chemosphere 38, 945–962.

Vallius, H. 1999b. Recent sediments of the Gulf of Fin-land: an environment affected by the accumulationof heavy metals. Åbo Akademi University. 111 p.

Vallius, H., & Lehto, O. 1998. The distribution ofsome heavy metals and arsenic in recent sediments

from the eastern Gulf of Finland. Applied Geoche-mistry 13, 369–377.

Vallius, H. & Leivuori, M. 1999. The distribution ofheavy metals and arsenic in recent sediments of theGulf of Finland. Boreal Environmental Research 4,19–29.

Vallius, H. & Leivuori, M. 2003. Classification of hea-vy metal contaminated sediments in the Gulf of Fin-land. Baltica 16, 3–12.

Verta, M., Ahtiainen, J., Hämäläinen, H., Jussila, H.,Järvinen, O., Kiviranta, H., Korhonen, M., Kuk-konen, J., Lehtoranta, J., Lyytikäinen, M., Mal-ve, O., Mikkelson, P., Moisio, V., Niemi, A., Paa-sivirta, J., Palm, H., Porvari, P., Rantalainen,A.-L., Salo, S., Vartiainen, T., & Vuori, K.-M.1999a. Organochlorine compounds and heavy me-tals in the sediment of River Kymijoki; Occurrence,transport, impacts and health risks. Organokloori-yhdisteet ja raskasmetallit Kymijoen sedimentissä:esiintyminen, kulkeutuminen, vaikutukset ja ter-veysriskit. The Finnish Environment 334. 73 p. (InFinnish)

Verta, M., Korhonen, M., Lehtoranta, J., Salo, S.,Vartiainen, T., Kiviranta, H., Kukkonen, J., Hä-mäläinen, H., Mikkelson, P., & Palm, H. 1999b.Ecotoxicological and health effects caused byPCP’s, PCDE’s, PCDD’s, and PCDF’s in River Ky-mijoki sediments, South-Eastern Finland. Organo-halogen Compounds 43, 239–242.

Verta, M., Lehtoranta, J., Salo, S., Korhonen, M., &Kiviranta, H. 1999c. High concentrations ofPCDD’s and PCDF’s in River Kymijoki sediments,South-Eastern Finland, caused by wood preservativeKy-5. Organohalogen Compounds 43, 261–264.

WGMS 2003. Report of the Working Group on MarineSediments in Relation to Pollution. ICES CM 2003/E:04.

distribution of heavy metals and arsenic in soft...

Documents